Professional Documents
Culture Documents
Introduction
2 3
3 4 4 6 6
7
7 8 8 9
10
10 11
12
12 12 12 13
5. Conclusion
15
References
16
Introduction
Fire is a disaster for those involved. A major concern is the high damage potential and increasing insurance premiums from fires. In this document we address the role of insulating material in the fire safety of buildings, with a special focus on EPS. We will show that in a properly designed and constructed building, insulation material plays only a minor role in fire safety. On the other hand, insulating material contributes enormously to energy savings for the heating and cooling of buildings. This is not only a financial saving but also a contribution to the mitigation of carbon dioxide emissions and the prevention of global warming. The unique properties of EPS make it the ideal, sustainable insulating material for many applications.
The purpose of this document is to clarify the fire performance of expanded polystyrene foam (EPS) when used as insulation material. It provides an overview of the facts on fire safe constructions using EPS building products. It is targeted as a reference to all interested parties, such as: building owners, architects, constructors, firemen, insurers, risk managers and risk engineers. For the members of EUMEPS, the central issue is understanding and addressing the interests of the people involved, whether this is the owner who wants to have a comfortable, healthy, safe and affordable home; or a construction worker who wants to have a reliable, sound and fail-safe product; or a fireman who wants to limit the risks he faces when he helps people in case of emergency.
above 45C are uncomfortable; a prolonged stay in a temperature above 65C can cause damage to the lungs and people cannot survive for long if temperatures are higher. After the occurrence of a flash over, the fire reaches its full size and further development is limited by the availability of oxygen through ventilation. After flash over the chances of saving people or the content of a room are minimal because of the temperature, lack of oxygen and damage to materials by heat and soot. Left to burn, a fire will eventually decay due to the lack of flammable material. EPS starts to soften at a temperature of about 100C, a temperature at which people have minimal chance to survive. At this phase of a fire there is hardly any oxygen left and the air is toxic because of high levels of carbon dioxide and carbon monoxide. During the phase of rapid fire development, the flash over, wood will self ignite at a temperature of about 340C and EPS at a temperature of about
Compartment temperature
Stage 2 Developing re
Decay stage
Ignition
Flashover
Overview of casualties and damage per region. [ref 2, 3] 450C. So the time to save people and material is limited to the first stage of a fire and this is independent of the insulation material. After a flash over people inside cannot be saved and the material value of the compartment will probably be a total loss. From flash over damage control only can be achieved by isolating the possible fire. EPS has a limited role in the design of fire resistant constructions used to compartmentalise buildings. EPS should only be applied in such constructions in combination with other fire resistant materials which perform the fire resistant role. tiated in 1988, in which performance based criteria play a central role. Some older regulation still makes descriptive statements. An example of this could be requirements for non-combustibility of insulation material. The performance based alternative is to have fire performance criteria for building elements like the floor, the wall, the ceiling or the roof. The performance based approach tends to result in improved fire safety at lower cost. This can be illustrated in the Netherlands and New Zealand where regulation is mainly performance based. The death rate caused by fire in the Netherlands is now 6,4 per million inhabitants and 9,6 in New Zealand, compared to 13,3 per million in Europe and as high as 25,0 in the USA, which has mainly descriptive based regulation. In addition, statistics indicate that performance based regulation is an effective approach for limiting fire damage.Damage caused by fire in the Netherlands is 0,2% of GNP and as low as 0,11% GNP in New Zealand compared to the European average of 0,27%. The costs of prevention in the Netherlands amount to 0,3% of the GNP and 0,18% in New Zealand. A country like Denmark, with a mainly descriptive based fire regulation, spends 60% more on fire prevention but has 95% more damage and 128% more deaths from fires than the Netherlands, with its performance based approach. [ref 2 and 3]
Reasons for high amounts of damage Insufficient fire prevention measures Increase in business continuity damage caused by a concentration of production facilities and supplies More expensive, yet vulnerable, production facilities Lighter but at the same time bigger and more complex buildings Larger fire compartments Failing fire compartmentalisation measures and fire doors High fire load Insurance and claim behavior: lower own risk and wider coverage Non-compliance of regulations in force
Make compartments! Take into account the size of the compartment as well as the value of the contents of the compartment and the importance of it for business continuity. An example of this could be to divide the production from the storage of goods. Regularly check if compartmentalisation measures are functional. The risk is that openings will be made in compartment walls (e.g. for ventilation pipes or electricity cable gutters) or that fire doors wont close anymore. Professional workmanship Good design and detailing is a first step, but professional workmanship is needed to ensure performance. Bad preparation, incorrect materials and poor execution of the work is the source of many problems. Reduce the fire load! The fire load of a building is made up of two components: the static and the variable fire load. The fire load of the building products used for the construction is called the static fire load. Normally, the most important factor is the variable fire load, consisting of the building contents. To reduce the fire load the first two items to review are the contents of a building and the surface materials within a room. Insulating materials are normally covered by surface materials like gypsum, stone or steel and will only contribute to the fire after this surface material fails. By the time of this failure flash over has occurred and the compartment is a total loss. Make use of active fire prevention measures! A high percentage of fires is caused by arson, so not only smoke alarms and sprinklers but also burglar alarms, fencing and entrance protection systems need to be considered.
Details are important! Unless warnings from the roofing constructor they choose the cheaper detail. Result; the wooden substructure caught fire.
Prevent failure of fire doors! According to research by the global insurance company Factory Mutual, failing fire doors play a negative role in two-thirds of all fire damages. Compartmentalisation fails because fire doors are open, e.g. by wedges keeping the heavy doors open. Other preventive measures Maintenance of the electrical installation. Short circuits cause many fires and can effectively be detected by regular infrared thermographic checks. Maintain a policy of hot work permits. These permits normally include measures such as the availability of hand fire extinguishers, the availability of a mobile phone and checking of the area for signs of fire after one hour. Prevent the storage of combustible goods against the outer walls of a building. These stored goods often are subject of arson and can cause the whole building to be destroyed by fire.
Insulation material type is not the most important factor, e.g. 2008 fire at a Gamma DIY store (NL) with non-combustible insulation.
Regulatory demands vary from country to country but in many cases the reaction to fire performance of the naked product is just a formal mandatory criterion. Where regulation is dominantly performance based, as intended by the CPD, requirements are based on building or construction elements. Recent European developments address this point of view and make it possible to perform reaction to fire tests on standardised build-ups, simulating end use applications. Producers can then declare the classification on the reaction to fire, simulating end use applications, on the product labels just outside the formal CE-box. Research by EUMEPS points out that the classification for reaction to fire for EPS in the standardised build-up behind gypsum is Euroclass B-s1d0. The same classification results for EPS behind profiled steel, which uses a standardised build-up simulating the end use of EPS in a flat roof construction with profiled steel. In both cases, this results in the same classification as the identical construction with mineral wool or PIR insulation.
1.5 CE-marking
Since May 2003 CE-marking of insulation products is mandatory under the Construction Products Directive (CPD). CE-marking can be seen as the passport for the free trade of building products within the European Union. Part of the CE-mark is the declaration of the reaction to fire classification of the product. This classification applies to the naked product as placed on the market. For naked EPS this classification is Euroclass D or E in case of fire retarded material and Euroclass F in case of non-fire retarded material (which is often used for packaging). In fact, this classification tells little about the fire performance of the building element in which the insulation product is used.
illustrated by a demonstration in which a hole is burned into a big block of EPS using a torch. As soon as the torch is taken away the fire extinguishes. The reaction to fire behaviour should be evaluated not on the material or product, but on the building element or construction element level (also called works). A basic design rule with EPS and other insulating materials is never to use the material uncovered. Because EPS never should be the material facing the fire, reaction to fire classification of the material or naked EPS product is only of formal importance. The layer really determining the reaction to fire behaviour is the surface layer of the construction, which faces the fire and covers the EPS insulation material. Using a combination of EPS insulation and specific cover layers it is always possible to design a construction which fulfils the fire requirements. Correctly applied and installed EPS is not of influence to the occurrence and the development of fire in a building. The excellent behaviour of EPS in constructions has been confirmed by recent studies by EUMEPS. Tested according to EN 13501-1 for the standardised buildups of EPS covered with gypsum and steel resulted in a B-s1d0 classification. The smoke part of this classification, the s1, is the best possible classification for a construction, which means there is little or no contribution to the production of smoke.
In the table below Prager [ref 8] shows that there is little difference in the contribution to the fire load from the various insulation materials if compared at an equal insulation value.
Thermal conductivity
Density
Heat of combustion
Fire load/m3
(W/mK)
0,035 0,040 0,045
(kg/m )
3
H (MJ/kg)
39,6 39,6 4,2
Qv (MJ/m )
3
20 32 170
92 169 107
In [ref 8] Prager shows the contribution to the fire load for a number of common insulation materials.
The toxicity of smoke fumes from EPS and various 'natural' materials.
Sample Smoke gases in a Fire EPS Carbon monoxide (standard Monostyrene grade) Other aromatic compounds Hydrogen bromide EPS-SE Carbon monoxide (fire retardant Monostyrene grade) Other aromatic compounds Hydrogen bromide Deal Carbon monoxide Aromatic compounds Chip board Carbon monoxide Aromatic compounds Expanded cork Carbon monoxide Aromatic compounds Emitted fractions (v/v) in ppm at different temperatures 300C 400C 500C 600C 50* 200* 400* 1,000** 200 300 500 50 fractions 10 30 10 0 0 0 0 10* 50* 500* 1,000* 50 100 500 50 fractions 20 20 10 10 15 13 11 400* 6,000** 12,000** 15,000** 300 14,000** 24,000** 59,000** 69,000* fractions 300 300 1,000 1,000* 3,000** 15,000** 29,000** fractions 200 1,000 1,000
Remarks: Test conditions specified in DIN 53 436; air flow rate 100 1/h; 300mm x 15mm 20mm test specimens compared at normal end-use conditions. * smouldering/glowing ** as a flame - not detected APME research according to DIN-53436.
Extensive research by APME, performed according to DIN-53436, at temperatures from 330C to 600C also proved that fire retarded EPS produces less toxic fumes than natural materials, producing no gases such as chlorine or cyanide [ref 11]. EPS combustion is relatively clean.
phase of a fire. Tested according to EN 13501-1 many applications will achieve a B-s1d0 classification. The s1 classification for smoke production is the best possible classification. When directly exposed EPS burns it produces a considerable amount of heavy, black smoke, which is proportional to the consumed mass. Hence, when used correctly in recommended applications, EPS does not contribute to the spread of fire and produces little smoke and toxic gases. The choice of the insulation material has little influence on the amount of toxic gases and smoke produced during a fire.
Compartments All buildings contained some kind of fire compartments, but only in 62% of the cases was this known to the fire fighters, who could then adapt their firefighting tactics accordingly. In 30% of the cases the compartmentalisation failed, 50% of which was due to failure of the self-closing fire doors. Time of the start of the fire Most fires started outside the normal opening hours of the building: between 18.00 hrs and 09.00 hrs. Extinguishing of the fire The fire fighters arrived at the fire within the acceptable time span after the fire was reported. In about 5% of the cases there was a problem to reach the fire and in 5% of the cases there was a problem with the availability of water to extinguish the fire. In 13% of the cases the fire fighters were not able to prevent the spread of fire to the neighbouring location. The fire fighters initially tried to fight the fire from the inside of the building in two thirds of the cases. Cause of fire Many fires were caused by malfunctioning or misused equipment (26%) or arson (23%). In reality the percentages of both causes are probably much higher, because the cause remains unknown for 40% of the fires.
11
rial applied. Tests commissioned by EUMEPS illustrate that a wall construction with only 9 mm gypsum has a classification of B-s1d0 [ref 21]. Normally, no tests are required for a cavity wall construction with an inner wall made out of stone [ref 20]. Tests performed by Austrian testing institutes as well as the firepolice and the Austrian fire fighters of Graz proved that EPS for ETICS also performs excellently. ETICS can achieve a reaction to fire classification of B-s1d0 and a full scale tests confirm these results [ref 25]. Extensive statistical research on 175 fires by the Polish fire fighter organisation pointed out that the occurrence of fires with ETICS using EPS was proportional to the market share of EPS and similarly for mineral wool [ref 26].
12
A report by the Association of British Insurers (ABI) acknowledges that in the case of buildings for the food industry or coldstores, foam plastic cores are to be preferred to other solutions due to hygienic reasons. They also comment that sandwich panels do not start a fire on their own and, with appropriate fire safety management, risks associated with the food industry can be controlled acceptably. Around hot work areas (frying pans, etc.) special measures are needed and precautions should be taken where electric cables go through these panels, because the metal skins can cut through the electrical insulation of the cables (independent of the type of insulation!).
The key conclusions for the fire behaviour of EPS steel sandwich panels are: Independent of the core material, all steel sandwich panels with a plastisol coating will have the same Euroclass: B. Comparative research shows the results of the SBI tests are fully in line with the larger and therefore more expensive room corner test, ISO 9705 [ref 19]. The differences in the test results of steel sandwich panels with an EPS core are minimal when compared to other core materials. The joint detail and the mounting and fixing details of the sandwich panel are very important to the result of the fire tests.
branes, where there is a considerable risk of fire [ref 27]. Hence, it is not the insulation material that is the main concern but the hot work combined with the risk of details. Both can and will be solved by the industry to make the flat roof a safer place. The European classification system for external fire, EN 13501-5, refers to four different methods mentioned in the ENV 1187. For each of these methods it is easily possible to design a construction with EPS insulation which fulfils the requirements. Normally, there is a layer of glass fleece included somewhere in the build up. The testing for the roof construction is usually commissioned by the producer of the roofing felt. Nearly all current waterproofing felts have been tested in combination with EPS because the roofing felt producer wants to make use of the superior qualities of EPS as a flat roof insulation material regarding durability, walkability, ageing and price. Many modern industrial buildings are made of a light weight steel construction. Sometimes the fire safety of this kind of building is a subject of discussion and the insulation material becomes part of the debate. In reality, the objective is for as large a building for as little money as possible and so cost as opposed to fire safety is the driving factor. A steel construction without any protective coating fulfills this criterion. If a fire starts in a compartment of such a building and can grow into a developed fire, then this part of the building is a total loss. Within 10 to 20 minutes the steel construction can collapse and fire fighters will not be able to enter the building. What is the role of the insulation material in this scenario? The true answer is that it has a fairly unimportant role.
13
Research has been commissioned by the EPS industry to find out the behaviour of different insulating materials in such a light weight steel construction [ref 12]. The conclusion of this research is that for EPS the time until a fire spreads from within the building to the surface of the roof is about 20 minutes. For other insulating materials this time could possibly be extended by another 10 to 20 minutes. It is questionable whether this is relevant if the construction fails normally between 10-20 minutes, before the fire spreads through the roof. Furthermore if a roof is not fully designed according to proven fire resistance, not all details will be fire resistant. Practical experience shows that fire will not spread to the roof through the construction but by details such as a roof light, a water drain, a ventilation pipe, a window in the wall, etc. Once the fire is on the roof fire incident reports show that the fire canwspread with a speed of up to 4 m/min depending on the weather conditions. The fact that EPS insulation is thermoplastic has positive side effects in case of fire. The EPS shrinks away from the heat, returning to its original solid granular form and by doing so it loses its insulating properties. Therefore, part of the heat produced by the fire gets out through the roof. Because of this, the time to flash over is longer and the time before the steel structure collapses is extended. Hence, the firefighters will have more time to protect neighbouring compartments [ref 12]. A factor often not included in the analysis of fire behavior of constructions is the influence of moisture barriers and anti-corrosive coatings. Bituminous moisture barriers are still often recommended because they are the most effective and reliable moisture barriers. Other light weight moisture barriers such as PE film can easily blow away and tear. Unreliable moisture barriers can lead to severe problems in flat roofs
such as loss of insulation value by moisture saturation, loss of compressive strength and leaking due to the heads of the mechanical fasteners puncturing the roof when it is walked upon. Moisture barriers as well as anti-corrosive coatings heavily influence the reaction to fire behavior of the roof. Finally, a factor often overlooked is that a thermoplastic insulation material may shrink away and become useless after a severe fire, but other insulating materials have to be fully replaced as well. Why? - because you cannot eliminate the impregnated smell of the smoke. You can ventilate for ages but it wont help! So, in conclusion, the insulating material does not play a decisive role in the development of a fire in a lightweight building with a steel deck. If a fire starts within a compartment of such a building, this compartment is generally a total loss, if not for the fire then for the smoke and the lingering acrid smell.
14
Building design is important in order to find the right balance between the advantages and disadvantages of big compartments. On the one hand bigger compartments are cheaper to build with logistic advantages, but they have higher risks and higher insurance premiums. On the other hand smaller compartments are more inconvenient and the costs of prevention higher . Compartmentisation is key to management of the fire risk. Proven constructions and details need to be used, to maximise fire and smoke resistance. Design is important but attention is also needed for the construction and the maintenance phase. Recent research commissioned by EUMEPS, carried out by TNO/Efectis and by Warrington Fire Gent, concerned the reaction to fire of EPS on a steeldeck
according to EN 13501-1. This resulted in a Euroclass B-s1d0 classification. Despite this d0 classification, which is the best possible classification as regards the forming of burning droplets, questions still arise as to the possibility of molten EPS droplets falling down through the joints of a steel deck during a fire. Could such drops lead to the further spread of fire? If fire retarded EPS is exposed to fire it will shrink away. If further heated it will melt and droplets can fall down. However, these droplets extinguish as soon as they touch the ground and cool down. Tests prove that even tissue paper will not ignite by these droplets. If the droplets fall into an area already on fire they will not cool down and they will burn. The chance that a fire fighter or another person is hurt by EPS droplets is small.
5 Conclusion
Fire safety is one of the essential requirements when designing a building. It cannot be compromised.The role of insulation in respect to fire safety is often overrated. This document shows that it is perfectly possible to design a building using EPS as insulation material and fulfill all insulation requirements including fire safety.
15
REFErENCE LIST
[1] I nternational Standardisation Organisation (ISO), Technical Report 9122-1 [2] 3231, World fire Statistics, GAIN, nr 19, 2003 [3] 3232, VIB, Aktuelle Brandschutzkonzepte, Schneider e.a., TU Wien, april 2000 [4] 3157, ROOFS ,De vuurbelasting van een dak, Appels, Chr., september 2002 [5] 3230, Impact on Insurance, Battrick,P. FM Global, presentatie oktober 2001 Luxemburg [6] 3172, ASPO presentatie 26-01-2001, Las, H.E. [7] 3204, EUMEPS APME TR 01/2000 testing naked EPS, november 2000 [8] 2839, Research in the causes of fire, Prager, F.H., Cellular Polymers nr. 20-3 / 2001 [9] 3184, Omzetting Euroklassen, Mierlo, R. van, TNO, augustus 2001 [10] 2719, Long term fire behaviour of EPS B1 and B2, APME TD 99/01, februari 1999 [11] 3167, Fire behaviour of EPS, APME september 2002 [12] 0110, Brandgedrag gesoleerde stalen daken, TNO, Zorgman, H. , februari 1987 [13] 0514, Giftigheid van gassen bij verbranding EPS, Zorgman, H., TNO, juni 1980 [14] 2010 t/m 2013, Rookproductie EPS 15/20, -N/-SE, TNO, januari 1998 [15] 2798 t/m 2959, casustiek I, BDA, 2001-2002 [16] 3055, TNO, o.a. 2004/CVB-B0336/RNP/TNL [17] 3210, TNO, o.a. 2004/CVB-B0833/NSI/TNL [18] 3414, 2004 TNO-CVB-R0310 [19] 3189, Euroclasses of EPS/Gypsum, doublage, APME/EUMEPS, september 2004 [20] 2965, Onderzoek sandwichpanelen, Langstraat, W., TNO, maart 2002 [21] 2966, 2001 TNO-CVB-B04432 [22] 3166, ABI, Fire performance of sandwich panels [23] TNO rapport 2004-CVB-R0076, Paap,F., maart 2004 [24] 0857, Bevordering brandveilig werken, BDA/SBR rapport, november 1990 [25] Grossbrandversuch der Grazer Feuerwehr, september 2007 [26] Analysis of the response of thermal insulation to fire, fire hazard identification office, Poland, march 2004 [27] NVN6050 Eisen aan ontwerp en detaillering voor brandveilig werken aan daken, september 2006
16
19
05-08/1